基于TCN和双重注意力的股价预测模型

Stock Price Prediction Model Based on Dual Attention And Temporal Convolutional Network

股票市场的行情受许多变量因子的影响,目前针对时间序列的预测模型,往往难以捕捉多因子之间的复杂规律。针对上述问题,提出一种基于双重注意力机制和时间卷积网络(TCN)的股价预测模型。首先,使用一种更适合时间序列的卷积网络作为特征提取层,并引入特征注意力,动态挖掘输入的因子特征与收盘价的潜在相关性,其次,在门控循环单元(GRU)的基础上,引入时间注意力机制来提高模型学习重要时间点的能力,并从时间角度获得重要性量度。实验结果表明,所提出模型对于股价预测的误差指标相比传统预测模型表现更加优秀,且在指标特征和时间两个方面上实现了模型的可解释性。

The stock market is affected by many variables and factors,and the current forecasting models for time series are often difficult to capture the complex laws among multiple factors.Aiming at this problem,a stock price pre⁃diction model based on dual attention mechanism and temporal convolutional network(TCN)is proposed.First,a con⁃volution network that is more suitable for time series is used as the feature extraction layer,and feature attention is in⁃troduced to dynamically mine the potential correlation between the input factor features and closing prices,and sec⁃ond,on the basis of Gated Recurrent Unit(GRU)on the other hand,a temporal attention mechanism is introduced to improve the model's ability to learn important time points and obtain importance measures from a temporal perspective.The experimental results show that the proposed model performs better than the traditional prediction model in the error index of stock price prediction,and realizes the interpretability of the model in terms of index char⁃acteristics and time.